Fe doping effects on the structural, magnetic, and magnetocaloric properties of nano-sized Pr0.6Bi0.4Mn1−xFexO3 (0.1 ≤ x ≤ 0.3) manganites

Preparation of samples

The Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ (0.1 ≤  x  ≤ 0.3) samples were prepared by a modified sol–gel method [ ⁸ , ⁹ ]. Mixtures of Pr ₆ O ₁₁ , MnO ₂ , Bi ₂ O ₃ , and Fe ₂ O ₃ with up to 99.9 % purity in stoichiometric proportions are used in the desired proportion according to the following reaction:

The precursors were dissolved in dilute 1 M HNO ₃ . Afterward, 3 g of citric acid and 1 mL of ethylene glycol were added and a complete homogenous transparent solution was achieved. Then, the solution was heated on thermal plate under constant stirring at 80 °C to eliminate the excess of water and to obtain a viscous gel [ ¹⁰ ]. Finally, the powders were grinded in an agate mortar and annealed at 900 °C for 4 h [ ¹¹ ].

Characterization techniques

Synthesis

During the synthesis, the sample is an uncolored sol, then the gel turns yellowish, and after precipitation, pulverulent dark brown powders are obtained. After sintering at 900 °C, any significant modification is observed: the samples are pulverulent dark brown powders. Yield ranges from 88, 89, to 92 %, respectively, for x  = 0.1, x  = 0.2 and x  = 0.3, pointing out the efficiency of the sol–gel method.

FTIR spectroscopic investigations

Among numerous applications of vibrational spectroscopy to solid-state problems, those dealing with structural evolution of samples obtained by sol–gel process appear particularly interesting. IR spectroscopy gives qualitative information about the way in which the adsorbed molecules are bounded to the surface as well as structural information of solids. Figure  ¹ shows FTIR spectra of the different Pr _0.6 Bi _0.4 Fe _x Mn _1-x O ₃ systems.

Fig. 1

Figure  ¹ a shows an example of the as-prepared sample for x  = 0.3. The absorption bands in the 4000–1000 cm ⁻¹ range are characteristic of –OH, C=O, N–H, and CH ₃ functional organic groups, while the region between 1000 and 400 cm ⁻¹ is the fingerprint of the inorganic material. The observed broad bands at 3400 and 2320 cm ⁻¹ can be attributed to the O–H stretching vibration of water molecules [ ¹² – ¹⁴ ]. The bands at 1500 and 1066 cm ⁻¹ are ascribed to the stretching vibrations of the aliphatic CH ₃ groups [ ¹⁵ ]. The bands between 900 and 400 cm ⁻¹ are mainly due to metal oxide bonds (Bi–O, Fe–O and Mn–O). Similar results are obtained for x  = 0.1 and x  = 0.2.

The FTIR spectra of the sintered samples displayed in Fig.  ¹ b show only two bands. The 960 cm ⁻¹ band is assigned to the vibrations of COOH. The observed broad peak in the low wavenumber range of 560 cm ⁻¹ is characteristic of the Mn–O group [ ¹⁶ ] and can be attributed to the MnO ₂ layers linked to the perovskite layers. When the sample was heat treated at 900 °C for 4 h, the absorption characteristic of νO–H stretching vanished, due to water evaporation from the surface of the powder. Our results are comparable to those obtained by Sana et al. [ ²¹ ] for Pr _0.6 La _0.1 Ca _0.3 Mn _{1− x} Fe _x O ₃ prepared by sol–gel method and Abir et al. [ ¹⁷ ] for Pr _0.6 Sr _0.4 Mn _{1− x} Fe _x O ₃ prepared by ball Milling and evidence the formation of the desired manganites.

Structural properties

Phase identification and structural analysis were carried out by XRD technique with Cu K α radiation at room temperature. The XRD patterns of Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ heated at 900 °C for 4 h (Fig.  ² ) show that all the samples crystallize in the orthorhombic structure with Pnma space group. The identification revealed the coexistence of two structures. The first one corresponding to PrMn ₂ O ₅ fitted in the orthorhombic structure with the Pbam s pace group [ ¹⁸ – ²⁰ ]. The second phase fits with Pbnm space group (JCPDS files 00-054 0871) and corresponds to the expected orthorhombic perovskite structure associated to the Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ [ ²¹ ]; the percentage of the secondary phase is nearly constant and is around 20 %. Precisely, the yield ranges from 22.5, 20.9 to 18.9 %, respectively, for x  = 0.1, x  = 0.2 and x  = 0.3.

Fig. 2

The average particle diameter, D , was obtained using Scherer’s formula [ ²² ] for the peak width broadening as a function of the size of the particles.

Transmission electron microscopy

In order to better evidence the particle size of Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ powders obtained by the sol–gel method, transmission electron microscopy analysis (TEM) was carried out. Figure  ³ presents, respectively, the TEM micrographs of the Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ powders heat treated at 900 °C for 4 h. Particles of few tens of nanometres to micrometre in size are observed.

Fig. 3

Energy-dispersive analysis of X-ray (EDX)

Figure  ⁴ shows the EDX spectra of all the sintered samples. It is clear from the EDX spectra that these samples are composed of Pr, Bi, Fe, and Mn elements. The presence of Cu is due to the grid support. The analysis of the amount of the different elements has been performed by EDX at room temperature. Regardless of the sample, the general composition is close to the expected Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ ( x  = 0.1, x  = 0.2 and x  = 0.3) formula. Local quantification highlights that for x  = 0.2 and 0.3, the composition corresponds to the Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ formula, while for x  = 0.1, local inhomogeneity is observed. Such results were confirmed by SEM measurement (see below). Of course, due to the presence of the PrMn ₂ O ₅ minority phase, the exact composition of the perovskite structure is erroneous.

Fig. 4

Scanning electron microscopy

The SEM images of the as-prepared and sintered samples are shown in Fig.  ⁵ a. SEM micrograph of the as-prepared sample reveals that regardless of the iron amount, the obtained Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ samples are porous and homogenous but with not a well-defined particle shape (see Fig.  ⁵ a for x  = 0.2). On the other hand, the SEM micrographs (Fig.  ⁵ b) of the sintered samples evidence particles of well-defined shape. Interestingly, some modification of the particle shape is observed. While particles with a hexagonal shape are observed for x  = 0.1, particles exhibiting cubic shape are observed for x  = 0.2 and x  = 0.3 (Fig.  ⁶ ).

Fig. 5

Fig. 6

Magnetic properties

A systematic investigation of magnetization with temperature of all samples has been undertaken at very low applied magnetic field of 0.01 T (Fig.  ⁷ ). The variation of the magnetization ( M ) versus temperature. ( T ) reveals that all samples exhibit a ferromagnetic–PM transition at Curie temperature. T _C —defined as the temperature at which dM/dT shows a maximum value. The obtained T _C values are found to be 40, 77, and 52 K for x  = 0.1, 0.2, and 0.3, respectively. The evolution of T _C can be explained by the increased number of Fe ³⁺ –O–Fe ³⁺ interactions which indicates a weakening of ferromagnetism. This weakening can be explained by the competition between DE interaction Mn ³⁺ –O–Mn ⁴⁺ and superexchange antiferromagnetic interactions, typically Mn ³⁺ –O–Mn ³⁺ and Mn ⁴⁺ –O–Mn ⁴⁺ . The substitution of Mn by Fe giving rise to an antiferromagnetic coupling between Mn and Fe ions that favours the superexchange mechanism [ ²³ ].

Fig. 7

The PM–FM transition is also highlighted using a linear extrapolation to the zero magnetization of the χ− ¹ (T) curves (Fig.  ⁷ ). The Curie temperature of the Pr _0.6 Bi _0.4 Fe _x Mn _{1− x} O ₃ powders is similar to those obtained by Troyanchuk et al. and by Chen et al. [ ²⁴ , ²⁵ ]. The magnetic properties of PrMn ₂ O ₅ sample reveal that the oxide has antiferromagnetic transition at low temperature giving a Neel temperature ( T _N ) at 100 K.

In the PM region ( T  ≥  T _c ), the magnetization curve is well fitted by Curie–Weiss law (see Fig.  ⁷ the inset). The temperature dependence of inverse magnetic susceptibility data in the 5–300 K temperature range follow a Curie–Weiss behavior above T c. The relation between χ and temperature T should follow the Curie–Weiss law in the PM region:

In order to gain a deeper understanding of the magnetic properties and to confirm the ferromagnetic behavior at low temperatures, the isothermal magnetic hysteresis loops, M ( H ), have been recorded. As seen in Fig.  ⁸ , the curve reveals a ferromagnetic property with a clear magnetic hysteresis at 2 K. From this figure, we see that magnetization decreases with the increasing Fe content. In fact, magnetization does not reach the saturation for x  = 0.2 and x  = 0.3. This confirms the presence of the AFM interactions, which are more important for higher Fe contents ( x  = 0.2 and x  = 0.3) [ ²⁷ , ²⁸ ].

Fig. 8

Magnetocaloric study

The MCE is an intrinsic property of magnetic materials. It is the response of the material to the application or removal of magnetic field, which is maximized when the material is near its magnetic ordering temperature (Curie temperature T _C ). The evolution of magnetization ( M ) versus magnetic field ( H ) for the x  = 0.1 and 0.3 composition, obtained at different temperatures ( T ), is shown in Fig.  ⁹ . These curves reveal a strong variation of magnetization around the Curie temperature indicating a possible large magnetic entropy change associated with the ferromagnetic–PM transition temperature.

Fig. 9

In order to check the nature of the magnetic transitions for the nanopowder, we have used the criterion given by Banerjee [ ²⁹ ]. This criterion consists in inspecting the slope of isotherm plots of M ² versus H / M . From a thermodynamic point of view, it can be concluded that if all the curves have a positive slope, the magnetic transition is of the second-order type. On the other hand, if some of the curves show a negative slope at some point, the transition is of the first-order type. A more standard type of Arrott plots is observed. The plots exhibit a linear behavior, and the positive and negative slopes are clearly noted in the complete M ² range. So, following Banerjee’s criterion, we can conclude that the ferromagnetic-to-PM (FM–PM) phase transition in the samples is obviously of the second-order type for x  = 0.1 and first-order type for x  = 0.3 as shown in Fig.  ⁹ (the inset).

The MCE of the materials can be derived from the Maxwell’s thermodynamic relationship:

The values of the magnetic entropy changes (Δ S _M ) upon application of magnetic fields from the isothermal M – H curve can be numerically calculated using Eq. ( ⁴ ):

Figure  ¹⁰ shows the temperature dependence of −Δ S _M ( T , (Δ µ ₀ H ) for x  = 0.1 and x  = 0.2 for different applied magnetic fields. We can note that the magnetic entropy change depends on the magnetic applied field.

Fig. 10

As expected, the overall value of entropy change is found to increase with the increasing field. Another notable feature is that increasing field shifts the magnetic transition temperature to higher temperatures. The maximum values of magnetic entropy change, Δ S _M are 0.47 and 0.37 J/kg/K for x  = 0.1 and x  = 0.2, respectively. But this field is not enough to saturate the sample. Compared with Gd 2.8 J/kg/K at 1 T field [ ³⁰ ], Δ S _max value is smaller for the present materials. However, these materials can be easily synthesized with good chemical stability. Another interesting feature in the MCE plot is that it is asymmetric, especially under high field. Similar behavior is observed in Si-substituted Lanthanum calcium manganite [ ³¹ ]. It is difficult to determine their RCP because of the non-uniform distribution of the Δ S _M curves. The non-uniform distribution of Δ S _M in nanocrystalline manganites is due to the enhanced grain boundary effects and the spread of the ferromagnetic transition temperature in different magnetic clusters caused by the inhomogeneity of the structure and stoichiometry [ ³¹ ].